1

Suppose we have three variables $y_t$, $x_{1t}$, $x_{2t}$. $y_t$ and $x_{1t}$ are co-integrated. $x_{2t}$ is stationary. Now, is the regression model $$\Delta y_t = C + \beta_0(y_{t-1}-\alpha x_{1,t-1}-c)+ \beta_1\Delta x_{1t} + \beta_2 x_{2t}+\epsilon_t$$ valid?

Richard Hardy
  • 54,375
  • 10
  • 95
  • 219
oopsroger
  • 11
  • 1
  • Is that a homework question? If so, please familiarize yourself with the [`self-study`](http://stats.stackexchange.com/tags/self-study/info) type of questions and add the [tag:self-study] tag. – Richard Hardy Nov 14 '16 at 07:38
  • 4
    See ["VAR or VECM for a mix of stationary and nonstationary variables"](http://stats.stackexchange.com/questions/148994/var-or-vecm-for-a-mix-of-stationary-and-nonstationary-variables/149263#149263) and ["Modelling a nonstationary variable with stationary and nonstatianary variables"](http://stats.stackexchange.com/questions/147582/modelling-a-nonstationary-variable-with-stationary-and-nonstatianary-variables/147595#147595). – Richard Hardy Nov 14 '16 at 14:17

0 Answers0